Impact of hydro-meteorological conditions and flash drought duration on post-flash drought recovery time patterns

Abstract. Recovery time, which refers to the duration an ecosystem needs to revert to its pre-drought state, is a fundamental aspect of ecological resilience. Recently, flash droughts (FDs) characterized by rapid onset and development have been gained recognition. Nevertheless, the spatiotemporal patterns of recovery time and the factors that affect it remain largely unknown. In this study, we set up a novel method to investigates the recovery time patterns of terrestrial ecosystem in China based on gross primary productivity (GPP) by employing the Random Forest (RF) regression model and the Shapley Additive Prediction (SHAP) method. A random forest regression model was developed for analysing the factors influencing recovery time and establish response function functions through partial correlation for typical flash drought recovery periods. Additionally, the dominant driving factors of recovery time determined by using the SHAP method. Results reveal an average recovery time of approximately 37.5 days across China, with central and southern regions experiencing the longest recovery time. Post-flash drought radiation emerges as the primary environmental factor, followed by aridity index and post-flash drought temperature, particularly in semi-arid/sub-humid areas. Temperature exhibits a non-monotonic relationship with recovery time; with excessively cold or overheated temperatures leading to longer recovery times. Herbaceous vegetation recovers more rapidly than woody forests, with deciduous broadleaf forests demonstrating the shortest recovery time. This study provides valuable insights into comprehensive water resource and ecosystem management, and it will be helpful in large-scale drought monitoring.